Technically, then, there are no fields either - claiming that fields are real because the field model matches experiments better, commits the same mistake of taking the map as real, as when we said that particles are really there. As in Kantian noumena, we can't know the territory, we can just build more useful models.